Attention-Aware Feature Aggregation for Real-time Stereo Matching on Edge Devices

This repository contains the code (in PyTorch) for "Attention-Aware Feature Aggregation for Real-time Stereo Matching on Edge Devices" paper (ACCV 2020) by Jia-Ren Chang, Pei-Chun Chang and Yong-Sheng Chen.

The codes mainly bring from PSMNet.

Citation

@InProceedings{Chang_2020_ACCV,
    author    = {Chang, Jia-Ren and Chang, Pei-Chun and Chen, Yong-Sheng},
    title     = {Attention-Aware Feature Aggregation for Real-time Stereo Matching on Edge Devices},
    booktitle = {Proceedings of the Asian Conference on Computer Vision (ACCV)},
    month     = {November},
    year      = {2020}
}

Train

As an example, use the following command to train a RTStereo on Scene Flow

python main.py --maxdisp 192 \
               --model RTStereoNet \
               --datapath (your scene flow data folder)\
               --epochs 10 \
               --loadmodel (optional)\
               --savemodel (path for saving model)

Pretrained Model

KITTI 2015 Pretrained Model Google Drive